I've seen people complain about this before and I wonder how I always manage to avoid it. One thing that confuses me is why would somebody need to repeatedly go back to edit a source sequence in a multicam project.
When I have a multicam edit, I have a sequence that is used to sync. I may Color Correct, Scale Adjust, Re-sync additional footage, and make audio adjustments. But after that the syncing sequence is done and the actual multi-cam editing is ready to begin. I would only go back to the Sync sequence if I had made a mistake somewhere.
As a time saving measure, I always copy all of the viable audio tracks from the sync sequence and paste it, synced into the multicam sequence. That way I always have the raw audio to work with.